Bacteria use protein‑guided DNA synthesis
- Stanford researchers reported in Science that a bacterial antiphage system called DRT3 makes defensive DNA with two reverse transcriptases and one ncRNA. - One enzyme copies an RNA ACACAC motif into poly(GT), while the other builds the matching poly(AC) strand from a protein template. - That matters because sequence-specific DNA synthesis now looks broader than textbook template copying, with new implications for bacterial immunity and synthetic biology.
Bacteria have a lot more immune tricks than the old textbook version lets on. This one is especially strange. A Stanford team just described a defense system called DRT3 that helps bacteria survive phage infection by manufacturing long stretches of double-stranded DNA — and half of that DNA is made without reading any DNA or RNA template at all. That is the real news here. It pushes DNA synthesis into a category that basically was not supposed to exist. ### What is DRT3, exactly? DRT3 is a bacterial antiphage defense system built from three parts — two reverse transcriptases called Drt3a and Drt3b, plus a noncoding RNA. Reverse transcriptases usually copy RNA into DNA. But defense-associated reverse transcriptases have been turning up as weird, specialized immune enzymes in bacteria, and DRT3 is one of the strangest yet. The Stanford group showed that these three components assemble into a large 6:6:6 complex. (science.org) ### What does it make? It makes long, repetitive double-stranded DNA with an alternating pattern: one strand is poly(GT), the other is poly(AC). That sounds simple, but the important part is that the sequence is not random junk. The system reliably produces a defined repeating product, which means the enzyme machinery is controlling sequence, not just dumping nucleotides into a chain. ### Why is that weird? (science.org) Because polymerases are supposed to follow one of two playbooks. Either they copy an existing nucleic-acid template, or they build something template-free but low-information — short motifs, homopolymers, messy tails. DRT3 breaks that split. One half of the system uses a normal-ish nucleic-acid template. The other half still makes a defined sequence, but without one. That is the conceptual jump. ### Which part follows the normal rules? Drt3a does. It reads a conserved ACACAC sequence embedded in the ncRNA and uses that as a blueprint to synthesize the poly(GT) strand. So far, weird bacterial immunity but still recognizable biochemistry. There is an RNA template, and the enzyme copies it into DNA. ### So what is the bizarre part? Drt3b makes the complementary poly(AC) strand in the complete absence of a nucleic-acid template. (science.org) The paper argues that Drt3b uses its own protein architecture to enforce the alternating pattern. Basically, the enzyme itself acts like a mold for which nucleotide comes next. If that holds up broadly, it means proteins can guide sequence-specific DNA synthesis more directly than biologists had appreciated. ### When does the system turn on? Not all the time. In the experiments, DRT3 protected engineered E. coli against phage, but activation depended on a phage protein called ST61 in phage T1. The team also isolated phage escape mutants to pin down that trigger. That matters because these systems are not just weird polymerases sitting around — they are sensors tied to infection. ### How does this fit the bigger pattern? (science.org) It fits a fast-growing idea that bacterial immunity is full of nucleic-acid chemistry we barely recognized. Other DRT systems already showed odd behaviors — one can make long poly(dA), another can generate concatemeric DNA that effectively encodes a new gene during infection. DRT3 adds another branch to that family tree: protein-guided synthesis of a defined DNA repeat. ### Why should anyone outside microbiology care? Because this is the kind of mechanism that later turns into a tool. CRISPR started as bacterial immunity too. DRT3 is not a ready-made platform yet, but it hints that programmable DNA synthesis may not need a conventional template in every case. That could matter for synthetic biology, molecular recording, and engineered bacterial defenses. For now, the big takeaway is simpler — bacteria are still teaching us that the basic rules of information storage and copying in cells are more flexible than we thought. (science.org 1) (science.org 2)